[z92]

His conclusion :

> In the years to come, industry experts expect low-level coding and programming jobs to become obsolete with the development of new tools that remove the need to code completely.

Me wondering how much of it shall we see.

[arcturus17]

It's been discussed ad nauseam here, but companies love "no-code" tools until they find their limitations (which is often sooner rather than later) and then they desperately get out their wads of cash to have the problems they caused sorted.

This creates an entire ecosystem of "low-code-real-developers" who have the skill, commercial foresight (and possibly patience) of working inside the guts of low-code tools, doing Java, JS or what-have-you, but inside an ugly, sparsely-documented propietary API instead of an open-source framework with literal gigabytes of documentation and knowledge accumulated in Stack Overflow over the course of fifteen years...

So now instead of two in-house analyst-developers building business applications, you have Mark from Marketing & Sales building apps in low-code, and four "low-code-high-code" contractors on expensive contracts cleaning up behind Mark.

And so it goes...

[nickthemagicman]

Haha this is the business model for every project I've ever worked on in Tech is.

Hire low paid developers to 'save money' at the start.

Low paid developers present working prototypes that aren't scalable or maintainable but the biz people don't know that ,they just see working demos.

Years later the app is a nightmare and the low paid devs have moved on and now they have to pay out the nose to get people to actually work on their stuff.

And so it goes.

[ryandrake]

> Years later the app is a nightmare and the low paid devs have moved on and now they have to pay out the nose to get people to actually work on their stuff.

You forgot the part where they keep trying to lowball competent developers, struggle to find someone willing to duct tape their system together for "competitive" pay, and throw their hands up declaring there's a Shortage Of Developers™.

[sanxiyn]

Being able to do something by oneself is such a huge advantage over telling someone else to do something that it can compensate for lots and lots of disadvantages. HyperCard was a great success, and I have high hope we will have great low-code tools like HyperCard in the future.

[arcturus17]

I was being facetious.

I agree there can be massive advantages to tools in the "lower-to-no-code" spectrum; even software developers benefit from the abstraction that frameworks and tooling provide. At a business level, many organizations that offer products and service in said spectrum have been successful at a massive scale: Wordpress, Salesforce, Tableau, Shopify, just off the top of my mind.

However my point about these applications creating a cascade of new jobs still stands. And I don't think more hand-crafted "high code" solutions are going anywhere either; for starters I don't think anyone making this sort of prediction understands the literal oceans handcrafted of code we swim in. Who's going to take care of that, and how long would it take to replace with "no-code"? Well over a century, would be my guess. Software hasn't eaten the world - it has devoured it.

Then there's the natural limitations of these tools. I had to look up HyperCard as I didn't know about it, and it was indeed an initial hit, but it seems like (a) it was limited (for example, Wikipedia mentions people were making "choose-your-own-adventure" games with it, but I know as a fact that some games made by solo devs in the late 80s where infinitely more intricate and deep) and (b) it didn't have that much lasting power. I wonder if the latter is yet another limitation of these tools - that they tend to quick obsolescence as tech trends move at breakneck speed.

[ebiester]

I will say, as much as I agree with you, even software organizations have problems that are well suited toward these low code platforms. After working through some of the quirks of retool, we have developers just adding things in the fraction of the time as they took with our first solution. I also saw another developer create an "excel++" solution with Airtable that allowed us to solve some problems that would have taken twice as long with twice as many people otherwise.

Now, this isn't every problem, but it's a very useful tool that will reduce the number of developers for some tasks. I think there is a world where a business has six or seven licenses for no-code solutions that reduce the head count by 10-15 FTEs (or redirects them)

[scroot]

> it didn't have that much lasting power. I wonder if the latter is yet another limitation of these tools - that they tend to quick obsolescence as tech trends move at breakneck speed.

Hypercard dying on the vine has much more to do with Apple's changing priorities than it does with Hypercard itself or what its users thought.

[rory]

The thing is, high-level languages like Python are already a really efficient way to communicate requirements to a computer, and we've already seen a massive shift from low-level languages to high-level languages in applications that don't need low-level control.

Developer tools and user interface components will likely continue to get easier to work with, but chunks of functionality that we compose with arbitrary code in between will always be a better model than a "no-code" monolith we configure.

[TheCoelacanth]

Yes, I think the doom-and-gloom predictions are dramatically over-hyped.

I'm pretty sure that software development has had by far the most effort put into automating it of any job that has ever existed, simply because every person capable of doing the job has the same skills necessary to automate it.

That has done nothing to decrease salaries for developers. If anything, it has raised them, simply because developers can now produce more useful things in less time and at lower overall cost.

[Nasrudith]

Personally I see lawyers as "proof" that taking the highest level won't be sufficient to do away with devs. That and many of the non-technical specifications emphasize that there must always be someone to specify the details and make sure they are consistent and handle the nasty corner cases.

[danjac]

Without context, I could guess that quote could have been written at any time in the past couple decades. It's the "nuclear fusion" of our industry: one day someone will make it happen, but not as soon as you think.

[durnygbur]

It's like instant on demand video and voice calls. Science ficiton had been dreaming of it since the end of 19th century. Then suddenly... Skype. It will happen in the most obscure place.

[RegBarclay]

As the developer tooling improves and parts of my job get easier, I sometimes think I may be out of a job someday. But then I talk to some users or other non-devs about technical issues and realize I will always have a job.

[after_care]

The delta in technical expertise between the average dev and the average user is so large that there will always be economically viable ways for the dev to spend their time.

[np-]

The way I see it, at the end of the day we still live in the real world with real, imperfect business/social/etc processes. Unless the entire world is 100% standardized and run by robots, any technological no-code solution would still need to be Turing Complete. At that point, basic computer science principles need to be learned to use it, so all you've done is create another programming language, which again...we already have plenty of those.

[visarga]

> Me wondering how much of it shall we see.

Have you heard of RPA? It's a low-code/no-code way to automate mouse and keyboard actions on top of desktop and web applications.

[rutthenut]

I think that sort of claim has been made for decades already, repeated by each generation of programming language, 4GL tools, AI and graphical tools.

[tonyedgecombe]

An example from forty years ago: https://en.wikipedia.org/wiki/The_Last_One_(software)